Introduction to R and Rstudio

Research Associate, Department of Sociology

This work is licensed under a Creative Commons Attribution 3.0 Unported License.

URL: https://compass-workshops.github.io/info/

• Teaching Staff

  • Ethan Fosse (Research Associate, Department of Sociology)
  • Yunkyu Sohn (Research Associate, Department of Politics)

• Faculty Sponsors

  • Margaret Frye (Assistant Professor, Department of Sociology)
  • Kosuke Imai (Professor, Department of Politics)
  • Matthew Salganik (Professor, Department of Sociology)

1. Workshop Requirements
2. Installing R and RStudio
3. Our Research Question
4. Roadmap for the Workshop

1. You have access to a laptop computer
2. You have access to reliable Internet service
3. You want to learn more about R and RStudio!
4. You have downloaded the R Workshop files:
   • MyFirstScript.R
   • States.RData
   • StatesHealth.dta
   • MyFirstMarkdown.Rmd

Go to: https://compass-workshops.github.io/info/

• R is free, open-source, and very popular!
• How are R and RStudio related?
  • R is the underlying programming language we use, while Rstudio provide a graphical front-end that has a number of useful features, including syntax highlighting and windowing
  • For this course we use RStudio with R, but there are other graphical front-ends that some people use with R

• Download R before installing RStudio
• Make sure to select the installer appropriate for operating system
• 1. Go to https://cran.r-project.org and follow the instructions for installing R
• 2. Go to https://www.rstudio.com/products/rstudio/download3 and follow the instructions for installing RStudio

• When you first open RStudio, you will be greeted by three panels:

1. The interactive R console (entire left)
2. Environment/History (tabbed in upper right)
3. Files/Plots/Packages/Help/Viewer (tabbed in lower right)

• Once you open files, such as R scripts, an editor panel will also open in the top left

• Throughout this workshop we will consider the following question:
  • Why did some states go for Obama and others for McCain?
• We'll explore this question using real data on 50 states
• In the process of answering this question we'll learn how to use R and RStudio for data analysis!

• Part 1: Exploring a Data Set
• Part 2: Working with Variables
• Part 3: Summarizing and Visualizing Data
• Part 4: Finding and Analyzing Subgroups
• Part 5: Importing, Saving, and Exporting Data
• Addendum: Creating Dynamic Reports

• Let's load States.RData into our workspace

• Using RStudio's user-friendly interface:

  1. File —> Open File
  2. Navigate to the location where you downloaded the files for this workshop (for example: C:/Folder/)

• You can also try this R Code:

  load(C:/Folder/States.RData)
  

• R-speak:
  • Data set is called a data frame
  • Text variables are called character vectors
  • Numerical variables are numeric vectors
  • Categorical variables are factors and their categories are called levels
• These are all called objects and appear in the Environment tab
• We can use the R function class() to see what we have!
• Try this R Code:

class(States)

• Let's look at our data in more depth!
• Let's use the function View()
• Try this R Code:

View(States)

• Ask yourself:
  • What are the observational units of the data set?
  • What variables look interesting or unusual to you?

• We can also view a data set in the R Console directly

• R functions: print(), head(), tail()

  • Try this R Code:

print(States)
head(States)
tail(States)

• Ask yourself:
  • What is the problem with the print() function?
  • What is the difference between head() and tail()?

• We can also examine the structure of a data set in a more compact way
• Let's use the function str()
• Try this R Code:

str(States)

• Ask yourself:
  • How many observational units are there?
  • How many variables are there?
  • Which variables are categorical? Numerical?

• The dimensions of a data set are the number of rows and number of columns
• R functions: dim(), nrow(), ncol()
• Try this R Code:

dim(States)
nrow(States)
ncol(States)

• Ask yourself:
  • What are these different functions doing?

• Besides the dimensions, it's often helpful to know the names of the observational units and variables
• R functions: rownames() and colnames()
• Try this R Code:

rownames(States)
colnames(States)

• Ask yourself:
  • Can you find Ohio in the set of row names?

• Did Obama or McCain win Ohio in 2008?
• R Code Hint:

View(States)

• Follow-up: Who won your home state?
  • Note: If you're not from the United States, then feel free to pick New Jersey!

• At this point you should have:
  • Loaded a data set into R
  • Viewed a data set as a spreadsheet
  • Explored the types of variables in a data set
  • Examined the dimensions of a data set

• The 1936 Election was a landslide (46 versus 2 states won):

• Was 2008 a landslide? How many states did Obama win versus McCain?

• Summarizing a data set using summary()

• Try this R Code:

summary(States)

• Ask yourself:
  • How many states are in the Northeast (NE)?
  • What is the average household income across states?
  • What is a drawback of using this function on a data set?

• We can grab a variable using the dollar sign ($) notation:
• General format: Dataset$Variable
• Try this R Code:

States$HouseholdIncome
States$Region

• We can use the summary() function on single variables!
• Try this R Code:

summary(States$HouseholdIncome)
summary(States$Region)

• We can use other functions we learned before as well, including class() and str()
• Try this R Code:

class(States$HouseholdIncome)
class(States$Region)

str(States$HouseholdIncome)
str(States$Region)

• Ask yourself:
  • Are these variables categorical or numerical?
  • What are the number of observational units for each variable?

• We can use the assignment operator (<-) to copy a data set
• General format: NewDataset <- Dataset
• Try this R Code:

StatesCopy <- States
str(StatesCopy)
class(StatesCopy)

• Ask yourself:
  • Is StatesCopy the same as States?

• We can also use “<-” to create a standalone variable
• General format: NewVariable <- Dataset$Variable
• Try this R Code:

HouseholdIncome <- States$HouseholdIncome
str(HouseholdIncome)
class(HouseholdIncome)

• Ask yourself:
  • Is HouseholdIncome the same as States$HouseholdIncome?

• Look at the Environment tab to see all the variables and data sets in the R workspace
• We can also show them in the R console: ls()
• How to remove a variable or data set: rm()
• Try this R Code:

ls()
Population <- States$Population
summary(Population)
rm(Population)

• Was 2008 a landslide? How many states did Obama win versus McCain?

• R Code Hint:

Winner <- States$ObamaMcCain
summary(Winner)

Or you can just do:

summary(States$ObamaMcCain)

• Follow-up: Can you delete the variable Winner?

• At this point you should have:
  • Summarized a data set
  • Copied a data set
  • Created standalone variables
  • Summarized and deleted standalone variables

• An election is more than just winning or losing
• Let's look at some other questions:

1. Across all states, what is the average percentage who voted for Obama?
2. How many states in the Northeast voted for Obama rather than McCain?

• Answering these questions requires using more R functions
• This can get laborious in the R console!

• In RStudio, go to File -> Open File and navigate to the location you downloaded MyFirstScript.R
• Right now your R script is empty!
• Type the following into the document:

nrow(States)
ncol(States)

• Highlight the lines of R code and click on Run or press Ctrl + Enter
• Cool, right?

• When conducting data analysis, R scripts can be really, really long
• So you and others can understand what you've written, add comments using the # symbol
• Add comments to your R script as follows:

nrow(States) # number of rows
ncol(States) # number of columns

• In Rstudio you can save your R script by clicking the Save icon or by pressing Ctrl + S
• Now you can open it up and run your code anytime!
• In RStudio you can also create new R scripts by going to: File -> New File -> R Script

• Just like writing prose or poetry, writing R code has certain conventions
• A few tips:
  • Always add a space after a comma
  • Add detailed comments whenever possible
  • Name objects (or variables) and files as clearly as possible
• For more see Google's R style guide
• Now let's analyze some data!

• Measures of central tendency
  • Mean: mean()
  • Median (or 50th percentile): median()
• Try this R Code:

mean(States$HouseholdIncome)
median(States$HouseholdIncome)

• Ask yourself:
  • Why is the mean different from the median?

• Measures of spread (or variablity)
• Range: range()
  • Standard deviation: sd()
  • Inter-quartile range: IQR()
• Try this R Code:

range(States$HouseholdIncome)
sd(States$HouseholdIncome)
IQR(States$HouseholdIncome)

• We can create a histogram with hist()
• Optional inputs (or arguments) in an R function are separated by commas
• We can control the number of bins with the option breaks
• Try this R Code:

hist(States$HouseholdIncome)
hist(States$HouseholdIncome, breaks=3)
hist(States$HouseholdIncome, breaks=15)

• In R-speak, categorical variables are called factors and the different categories they have are levels
• R function for examining levels: levels()

• To create a table of counts: table()

• Try this R Code:

levels(States$Region)
table(States$Region)

• Ask yourself:
  • How many states are in the West (W)?

• To create a table of counts for two variables we still use table()
• Now we just specify two variables as inputs
• Try this R Code:

table(States$Region,  States$ObamaMcCain)

• Ask yourself:
  • How many states are in the West (W) and voted for McCain?

• We can also create a bar plot of counts with plot()

• The height of the bars equals the number of observational units in each category (or level)

• Try this R Code:

  plot(States$ObamaMcCain)
  

1. Across all states, what is the average percentage who voted for Obama?
2. R Code Hint:

mean(States$ObamaVote)

1. How many states in the Northeast voted for Obama rather than McCain?
2. R Code Hint:

table(States$Region, States$ObamaMcCain)

• At this point you should have:
  • Summarized a variable based on its central tendency and spread
  • Created a histogram and scatter plot for numerical variables
  • Generated tables for categorical variables
  • Created a bar graph for a categorical variable

• The U.S. South is a distinctive cultural and economic region:

• What percentage voted for Obama in Mississippi compared to Massachusetts?

• The rows and columns are indexed by position and by name
• We can subset the data set using single bracket notation: [ , ]

• General format:

  • Subsetting to a row: data[row, ]
  • Subsetting to a column: data[ , column]
  • Subsetting to a row and a column: data[row, column]

• Ask yourself:

  • How many rows and columns are there in States.RData?

• General format for subsetting to a row: data[row, ]
• Try this R Code:

View(States) # look for the 1st row
States[1, ]
States["Alabama", ]

• Ask yourself:
  • What percent of people in Alabama voted for McCain?
  • What are other social and political characteristics of Alabama?

• General format for subsetting to a row: data[ , column]
• Try this R Code:

View(States) # look for the 2nd column
States[ , 2]
States[ , "HouseholdIncome"]

• Ask yourself:
  • Doesn't this look familiar?

• General format for subsetting to a row: data[row, column]
• Try this R Code:

View(States) # look for the 1st row and 2nd column
States[1, 2]
States["Alabama", "HouseholdIncome"]

• Ask yourself:
  • What is this result? Why is it a single number?

• At this point you should have:
  • Subsetted rows of a data set
  • Subsetted columns of a data set
  • Subsetted rows and columns of a data set
  • Combined elements with the concatenate function

• Our current data set does not have any variables on health
• But we do have StatesHealth.dta
  • This is a Stata data file (.dta extension)
• To load this data into R we will need to use the foreign() R package

• Since R is free and open-source, lots of people are writing R packages
• An R package is just a collection of R functions, data, and code in a well-defined format
• R packages can do everything from text mining (tm package) to data visualzation (ggplot2 package) to data wrangling (dplyr package)
• Over 9,000 R packages as of 2016 on CRAN (Comprehensive R Archive Network)

• To see all the R packages in CRAN:
• 1. Go to https://cran.r-project.org/web/packages
• 2. Click on one of these links:
  • Table of available packages, sorted by date of publication
  • Table of available packages, sorted by name
• I usually browse packages by date of publication

• To install R packages in RStudio: GUI versus R Console
• 1. Using the GUI: Go to the Packages tab and click Install
• 2. Using the R Console: Type install.packages("package_name")
  • package_name is just the name of the R package in quotes
  • Example: install.packages("foreign")

• Once you've installed an R package, it's then bundled with R and RStudio
• You now have access to all of the functions, data, code, and other files associated with the installed R package
• However, to access these files you must load your R package
  • Try this R Code: library(foreign)
• You only have to install an R package once, but you must load it every time you start R

• After importing a data set, it's a good idea to check it for possible errors
• Try this R Code:

View(StatesHealth)
head(StatesHealth)
tail(StatesHealth)

• Ask yourself:
  • How is this data set different from States.RData?

• To save the data set as an .RData file, we can use save()
• This will save the data set in the working directory
• We can verify that it's in the directory with dir()
• Try this R Code:

save(StatesHealth, file="StatesHealth.RData")
dir()

• You can also save all data, variables, and other objects in R with save.image()
  • Example: save.image("Everything.RData")

• Saving a data set will give it an .RData extension

• But we can the States data set as a Stata data set

• Try this R Code:

write.dta(States, file="States.dta")
dir()

• We can now open States.dta with Stata!

• R can import (and export) many kinds of data:
• Importing Excel Spreadsheets (.xlsx)
  • Package: xlsx
  • Function: read.xlsx()
• Importing SPSS Files (.sav)
  • Package: foreign
  • Function: read.spss()
• Importing SAS Files (.xpt)
  • Package: foreign
  • Function: read.xport()

• At this point you should have:
  • Imported a data set into R
  • Saved an R data set for later use
  • Exported a data set into other formats

• At this point you should have:
  • Loaded and explored the structure of a data set
  • Worked with an R script and installed an R package
  • Summarized and visualized numerical and categorical variables
  • Found and analyzed subgroups in a data set
  • Saved, exported, and imported a data set

• Please fill out this survey so we know how we can improve the workshop
• Link: https://goo.gl/forms/GbNl7fhCx70IRVpi1

URL: https://compass-workshops.github.io/info/

Email List: Send an email to listserv@lists.princeton.edu with “Subscribe COMPASSWORKSHOPS” in the body and all other lines blank, including the subject

• The R script is useful for replicating what you've you done in R
• But often you will want to create dynamic reports that combine text, code, and output
• In RStudio, go to File -> Open File and navigate to the location you downloaded MyFirstMarkdown.Rmd
• Three main components of R Markdown files:
  • R Instructions
  • R Code Chunks
  • Markdown Text

The initial chunk of text contains instructions for R

• Telling R the title, author, and date:

title: "My First Markdown"
author: "Ethan Fosse (COMPASS Workshops)"
date: "February 13, 2018"

• Instructing R to produce html output (in other words, a web page):

output: html_document

• The R Markdown file also has embedded R code chunks
• This is the R code chunk in MyFirstMarkdown.Rmd:

# loading the R data set
# load("C:/Folder/States.RData") 

# examining the data set
head(States)
nrow(States)
ncol(States)

• The chunk begins with 3 single back quotes followed an {r} and ends with 3 single back quotes

• The rest of the document consists of Markdown text
• Markdown is a system for writing web pages by marking up the text much as you would in an email rather than writing html code
• The marked-up text gets converted to html, replacing the marks with the proper html code
• You make text bold using double asterisks, like this: **bold**, and you make things italics by using single asterisks, like this: *italics*.

• Highlight the R code within the R code chunk and click Run or press Ctrl + Enter
• It works just like an R script!
• Now add the following to the R code chunk:

mean(States$ObamaVote)

• Re-run the R code chunk and see what happens!

• At the end of the R Markdown file type 3 single backquotes followed by {r}
• Type the following into the new R code chunk

ObamaMcCain <- States$ObamaMcCain
plot(ObamaMcCain)

• We also need to check that all of the R code chunks have the correct R code
• In the R Markdown file go to the following line of code:

# load("C:/Folder/States.RData")

• Remove the comment symbol # so that R will run this line of code
• Change load("C:/Folder/States.RData") to reflect the correct location of States.RData on your computer
• If you don't specify the correct location of the data set, then you will have an error when trying to create a report

• To create a report from your R Markdown file, click on Knit Html
• Click on the gear icon for viewing options:
  • View in Pane: Viewing the html file within RStudio
  • View in Window: Viewing the html file in a separate window
• Make sure your load() specifies the appropriate location
• Other output options are pdf and Word
• To create a new report, go to File -> New File -> R Markdown

• At this point you should have:
  • Learned about the format of an R Markdown file
  • Created and edited code chunks in R Markdown
  • Knitted an R Markdown file